nszd - GSI Environmental Inc.
Transcription
nszd - GSI Environmental Inc.
SOURC S CE REPORT D: NATUR N RAL SOURCE ZONE Z D DEPLET TION (N NSZD) PILOT P TEST T MEMOS M Is ssued: Octob ber 2015 Prepared by: The LA LNAPL Workgroup p NATURAL SOURCE ZONE DEPLETION (NSZD) PILOT TEST MEMOS Issued: October 2015 Prepared by: The LA LNAPL Workgroup COMMENTS ABOUT THIS DOCUMENT? Submit any comments or suggestions directly to: Mr. Mike Wang Western States Petroleum Association 970 W. 190th Street, Suite 770 Torrance, CA 90502 (310) 324-8565 TO CITE THESE DOCUMENTS: (1) Los Angeles LNAPL Working Group, 2015. Natural Source Zone Depletion (NSZD) Lines of Evidence and Results of Pilot Test at Shell Carson Facility. Western States Petroleum Association, Torrance, California. (2) Los Angeles LNAPL Working Group, 2015. Natural Source Zone Depletion (NSZD) Results of Pilot Test at Tesoro East Hynes Facility. Western States Petroleum Association, Torrance, California. AUTHORS (1) Poonam Kulkarni, Charles Newell, GSI Environmental Inc; Julio Zimbron, Colorado State University; Matt Himmelstein, AECOM (2) Poonam Kulkarni, Charles Newell, GSI Environmental Inc; Faraneh Chamran, Jun Lu, Daniel Buckley, Frank Muramoto, AECOM; Julio Zimbron, Colorado State University Memo 1: Natural Source Zone Depletion (NSZD) Lines of Evidence and Results of Pilot Test at Shell Carson Facility For Western States Petroleum Association October 2015 These memos were prepared by the staff of GSI Environmental, Inc., and AECOM on behalf of Western States Petroleum Association (WSPA) and the LA LNAPL Workgroup. The memos were prepared by/under the supervision of the engineers and geologists whose signatures appear below. This document is intended for the sole use and benefit of WSPA and its member industry companies. The scope of services performed as described in this memo may not be appropriate to satisfy the needs of other users, and any use or reuse of this document or the findings, conclusions, or recommendations presented herein is at the sole risk of said user. No express or implied representation or warranty is included or intended except that the work was performed within the limits prescribed by the Client with the customary thoroughness and competence of professionals working in the same area on similar projects. The findings or professional opinions were prepared in accordance with generally accepted professional engineering and geologic practice, and no warranty is expressed or implied. GSI Environmental, Inc. —ly i>' I Poonam R. Kulkarni, P.E. Environmental Engineer AECOM Matt S. Himmelstein, P.E. Principal Engineer October 2015 Page 1 of 22 MEMORANDUM TO: LA LNAPL Workgroup FROM: Poonam Kulkarni, Charles Newell, GSI Environmental Inc; Julio Zimbron, Colorado State University; Matt Himmelstein, AECOM RE: Natural Source Zone Depletion (NSZD) Lines of Evidence and Results of Pilot Test at Shell Carson Facility BACKGROUND The LA LNAPL Workgroup, working with Colorado State University (CSU), deployed carbon traps at the Shell Carson facility with the goal of estimating the carbon flux at the soil surface and by extension, estimating the rates of hydrocarbon biodegradation at the Shell Carson site. The lines of evidence supporting the use of carbon traps to estimate biodegradation rates and the results from the carbon trap pilot test at the Shell site are discussed below. LINES OF EVIDENCE CO2 is the product of hydrocarbon degradation. Thus, CO2 generation above background levels is direct evidence of the biodegradation of hydrocarbons, and part of a quantitative assessment of natural source zone depletion (NSZD) shown in the Figure 1 below (ITRC, 2009). In this analysis, we are measuring the “gas flux out” to determine if there is evidence for NSZD at the site. Figure 1: Example Control Volume “Box” for Quantitative Assessment of NSZD (Source: ITRC, 2009) October 2015 Page 2 of 22 Additionally, rather than collecting a CO2 profile in the vadose zone, as shown in Table 1 (i.e., highlighted box) and Figure 2 below (ITRC, 2009), we are using a new technique with the soil surface as a horizontal reference point. Table 1: Data for Qualitative Evaluation: Evidence that NSZD is Occurring (Source: ITRC, 2009) October 2015 Page 3 of 22 Figure 2: Reference Diagram for NSZD Mass Depletion Volatilization Process (Source: ITRC, 2009) Although capturing the CO2 flux at the soil surface is an emerging technique for demonstrating NSZD, the underlying principle is consistent with the National Research Council (NRC) approach, including the estimation of the CO2 fluxes generated by biodegradation. (NRC, 2000) “The decision to rely on natural attenuation and the confirmation that it continues to work depend on linking measurements from the site to a site model and “footprints” of the underlying mechanisms. Footprints generally are concentration changes in reactants (in addition to the contaminants) or products of the biogeochemical processes that transform or immobilize the contaminants.” October 2015 Page 4 of 22 (Source: ITRC, 2009). Finally, collection of CO2 gas as a biodegradation byproduct at the soil surface is less invasive than measuring a profile in the vadose zone, with easier implementation and lower costs. The CO2 measurements collected at the soil surface might not be fully representative of the CO2 profile in the vadose zone directly below due to potential drift caused by sub-surface heterogeneities, but several surface measurements across a site can be used to estimate the overall site CO2 flux. Additionally, CO2 fluxes at grade might also include CO2 naturally generated by the plant root zone, i.e., not related to petroleum biodegradation, that need to be accounted for and subtracted from the total carbon measurement. KEY POINTS CO2 is the product of hydrocarbon biodegradation. CO2 generation above background levels is direct evidence of biodegradation of hydrocarbons and an accepted indicator of NSZD. October 2015 Page 5 of 22 METHODS Two methods are presented below that can be used to estimate the hydrocarbon biodegradation rates at a site. Method 1: LNAPL Area vs. Background Area Method 1 involves the capturing and quantifying the CO2 flux at the soil surface at multiple locations across a site, but in two distinct areas: i) an LNAPL area (i.e. affected by hydrocarbons) and ii) a background area (presumed to have no LNAPL). By obtaining both types of values, the hydrocarbon biodegradation can be calculated as follows: Hydrocarbon Biodegradation Rate (gal/acre/yr) = Avg. Biodegradation Rate at LNAPL - Avg. Biodegradation Rate at Background Area (gal/acre/yr) Area (gal/acre/yr) The analysis can be complicated by heterogeneities in the subsurface. As seen in Figure 3 below, CO2 flux captured at the soil surface of a site with simple hydrogeology (i.e., mostly sandy soils) would be representative of the flux coming directly from a source zone. At a site with complex hydrogeology (i.e. layers of silts and clays), on the other hand, the CO2 emanating from the LNAPL body may migrate non-linearly and sideways, thereby establishing a wider CO2 footprint than predicted. The Shell Carson site has significant heterogeneities, so that some hydrocarbon biodegradation signal at the background sites was expected as part of this project. Figure 3: Potential CO2 Flux Pathways at Sites with Simple and Complex Hydrogeology Method 2: 14 C Analysis Method 2 relies on using the same carbon traps as in Method 1, but is based on the detection of 14C, which is a carbon isotope with a half-life of ~5,600 years. Petroleum hydrocarbon carbon sources are 14C poor, while “modern” (younger than 60,000 years) carbon sources are 14C rich. Thus, by quantifying 14C, it is possible to separate petroleum hydrocarbon carbon sources from more modern carbon sources. Effectively, October 2015 Page 6 of 22 this analysis allows for an automatic background correction that is required when measuring only CO2 flux from carbon traps. The 14C-based technique, thereby serves as a secondary method of determining biodegradation rates of hydrocarbons. Analytical costs are high for this type of analysis, so it is generally recommended to be used on a subset of samples (Zimbron, 2012), possibly as a confirmation of results from CO2 flux analysis. RESULTS A map showing the trap locations can be found in Attachment 1. Locations 5 and 10 were background locations at two fence line locations without hydrocarbon impacts (as indicated by no ROST signals in nearby borings), while the remaining traps were at LNAPL affected locations. Both CO2 and 14C analyses were run on all samples collected at Shell Carson (i.e. Methods 1 and 2). The results are shown in Table 2A and Table 2B below. The hydrocarbon biodegradation rate, as determined by applying Method 1, was found to be 1,700 gal/acre/yr, by subtracting the average background biodegradation rate (1,300 gal/acre/yr) from the average LNAPL affected area rate (3,000 gal/acre/yr). Consistent with the results from Method 1, the average hydrocarbon rate determined by Method 2 (14C analysis) was also 1,700 gal/acre/yr for the LNAPL locations. The 14C testing results also indicated that the “background” locations did have some hydrocarbon impacts. These hydrocarbon impacts at background locations are believed to be the result of either complex hydrogeology and the non-linear upward migration of CO2 from the source zone as previously described or another carbon dioxide source not associated with LNAPL Compared to hydrocarbon degradation rates at affected areas (average of 1,700 gal/acre/yr), however, the background area rates were much lower (average of 400 gal/acre/yr). There are two possible ways to evaluate background area rates in this method, depending on the source of the signal. If the measured carbon dioxide source comes from the LNAPL body, the average hydrocarbon degradation rate can be calculated to be 1,700 gal/acre/yr. If the carbon dioxide is coming from some other source not associated with LNAPL, then the measured hydrocarbon degradation rate at the background locations would have to be subtracted from the LNAPL locations, resulting in an average of 1,350 gal/acre/year (i.e., 1,700 gal/acre/yr – 350 gal/acre/yr). Since NSZD calculation Method 1 and Method 2 matched when it was assumed that the background hydrocarbon CO2 was coming from LNAPL, and because the actual source of the hydrocarbon-CO2 in the background cannot be confirmed, this analysis assumes that the background hydrocarbon CO2 source is the LNAPL body. This gives an average hydrocarbon degradation rate at the site of 1,700 gal/acre/yr using 14C analysis. October 2015 Page 7 of 22 Table 2A: Summary of CO2 Trap Results at Shell Carson Site (Method 1: Avg. from LNAPL Area Minus Avg. from Background Areas) Trap Number CO2 Trap Results (gal/acre/yr equiv.)* Location Description Affected Area Locations WIS-SH-1 West side of site WIS-SH-2 West side of site WIS-SH-3 West side of site WIS-SH-4 West side of site WIS-SH-6 Adjacent to biosparge pilot test WIS-SH-7 Adjacent to biosparge pilot test WIS-SH-8 North of pilot test WIS-SH-9 North of pilot test LNAPL Area Average Background Locations WIS-SH-5 South side of site WIS-SH-10 Northeast side of site Background Average = 1,500 5,100 1,900 2,725 2,000 4,200 4,100 1,800 3,000 = 800 1,700 1,300 Method 1: LNAPL Area (3,000) – Background Area (1,300) = 1,700 gal/acre/yr Table 2B: Summary of Analysis Results at Shell Carson Site (Method 2: 14C Analysis) Trap Number Location Description Affected Area Locations WIS-SH-1 West side of site WIS-SH-2 West side of site WIS-SH-3 West side of site WIS-SH-4 West side of site WIS-SH-6 Adjacent to biosparge pilot test WIS-SH-7 Adjacent to biosparge pilot test WIS-SH-8 North of pilot test WIS-SH-9 North of pilot test LNAPL Area Average Background Locations WIS-SH-5 South side of site WIS-SH-10 Northeast side of site Background Average Hydrocarbon Sources (gal/acre/yr equiv.)* 880 4,000 920 1,400 415 1,900 3,200 800 1,700 300 400 350** Modern Sources (gal/acre/yr equiv.)* Percent of Total Comprised of Hydrocarbon Sources (%) 660 1,200 960 1,300 1,500 2,300 900 1,000 1,200 57 78 49 53 21 45 78 44 53 480 1,300 900 39 24 32 Method 2: LNAPL Area Average Using 14C Analysis = 1,700 gal/acre/yr Notes: (1) Map of trap locations is provided in Attachment 1 (2) “Hydrocarbon” refers to LNAPL petroleum products in the subsurface, either in the vadose zone or saturated zone (3) LNAPL area and background area locations based on presence or absence of LNAPL (*) NSZD rates calculated using octane as the mineralization compound. See Question 3 (Key Questions) for more information. (**) See Method 1 description above for why hydrocarbon was detected in background areas. October 2015 Page 8 of 22 KEY POINTS The combined unsaturated/saturated zone hydrocarbon (LNAPL) biodegradation rate at the site averages approximately 1,700 gallons/acre/year equivalent. Two separate methods provided this same value: average for affected areas minus background (Method 1), and the average of the 14C-adjusted degradation rates (Method 2). NSZD is occurring at the Shell Carson site, and sources. 14 C analysis confirmed fossil fuel carbon KEY QUESTIONS Question 1. What are typical biodegradation rates (background and LNAPL areas) and how do they compare to those at the Shell Carson Site? Background CO2 flux rates for non-fossil fuel carbon sources can vary with climate and vegetation. Table 3A below shows background rates from McCoy, 2012. Background rates were measured at unimpacted areas of LNAPL-affected sites. (It should be noted that while these areas appear to be unimpacted based on site history and nearby ROST results, it is possible for these locations to have hydrocarbon impacts that are unknown). Here, the average background rate was determined to be 1,800 gal/acre/yr, which is similar to that of the Shell Carson site (1,300 gal/acre/yr). Table 3A: Background Results from Unimpacted Areas at Five Other Sites (Source: McCoy, 2012) Site A A A B B C C C C C D D E E E E E E E Sample ID CSU Background A-1 CSU Background A-2 CSU Background A-3 CSU Background B-1 CSU Background B-2 CSU Background C-1 CSU Background C-2 CSU Background C-3 CSU Background C-4 CSU Background C-5 CSU Background D-1 CSU Background D-2 CSU Background E-1 CSU Background E-2 CSU Background E-3 CSU Background E-4 CSU Background E-5 CSU Background E-6 CSU Background E-7 Gal/acre/year equivalent 850 1,800 950 1200 950 1,100 1,700 1,700 2,800 800 700 3,250 2,200 1,050 2,300 4,500 370 1,800 4,150 Avg. Background Degradation Rate = 1,800 Avg. Method 1 Background Rate for This Project: = 1,300 October 2015 Page 9 of 22 Notes: (1) Calculated gal/acre/year from CSU data in µmol/m2/sec using octane equivalent; however, measurements are assumed to represent non fossil fuel carbon sources and the gal/acre/year conversion is for comparison purposes only. Background rates were also determined using 14C analysis (see Table 3B) to identify the non-fossil fuel carbon sources in both LNAPL impacted areas, and those that were assumed to be unimpacted. The average background rate was determined to be 1,100 gal/acre/yr, which is similar to the background area average at Shell (1,200 gallons/acre/year). Table 3B: Background Results Based on 14C Analysis (Source: McCoy, 2012) Gal/acre/year Sample ID equivalent CSU Background A-1, C-14 697 CSU Background A-2, C-14 1,494 CSU LNAPL A-6,C-14 1,210 14 Avg. Background Based on C = 1,100 gal/acre/year Avg, Background Based on 14C for This Project (Method 2) = 1,200 gal/acre/yr Notes: (1) Applied % of non-fossil fuel carbon based on C14 analysis results from CSU to gal/acre/year calculation result at all locations The overall hydrocarbon biodegradation rates measured at the Shell Carson site are also comparable to rates measured by CSU at several other hydrocarbon sites as shown in Figure 4 below (McCoy, 2012). Here, LNAPL loss rates are synonymous with hydrocarbon biodegradation rates. October 2015 Page 10 of 22 Figure 4: LNAPL Loss Rates (gal/acre/yr) for 6 field sites (Source: McCoy, 2012) KEY POINTS Results from the pilot test background readings (average: 1,300 gal/acre/yr) are very comparable with background rates measured at other sites based on both readings at unimpacted locations (average: 1,800 gal/acre/yr) and 14C analysis of results (average: 1,100 gal/acre/yr). Hydrocarbon biodegradation rates from the pilot test (1,700 gal/acre/yr) are also comparable to those at 6 other field sites measured by CSU (McCoy, 2012), where the average LNAPL loss rate is 3,500 gal/acre/yr. Question 2. What Compounds in the LNAPL are Being Degraded? Site research studies with LNAPL have identified multiple attenuation processes that can degrade a wide range of hydrocarbons, from very light compounds (in the C4 and C5 range) to low volatility/lower solubility compounds up to the C30 range (e.g., Molins et al., 2010). Methanogenesis, for instance, exhibited opposite characteristics of the aerobic degradation sequence, in that it first depleted the >C18 alkanes (Bekins et al., 2005). Some LNAPL experts place NSZD and subsequent remediation technologies into one of two categories: October 2015 Page 11 of 22 i) Phase change - a process that enables preferential removal of certain compounds based on their physical, chemical, or biodegradation rates. Removal of volatile compounds via SVE is an example of phase change, as non-volatiles are not removed; and ii) Mass removal - a process that targets most of the compounds in a hydrocarbon mixture. Therefore, at early-stage gasoline release sites NSZD may act more like a "phase change" process, where the most soluble and volatile compounds are removed and degraded. At last-stage gasoline releases, NSZD will act more like a "mass removal" process, where methanogenesis slowly degrades long-chained alkanes. Due to the age of the releases at the Shell facility, NSZD may be acting more like a mass removal technology. Question 3. When Calculating Degradation Rates, Does Compound Selection Make a Difference? Carbon traps directly provide measurements of CO2 flux in µM/m2-sec. To make this value more meaningful, and because carbon dioxide is the end product of hydrocarbon biodegradation, it is often converted to an equivalent degradation rate in gal/acre/yr, by using a compound to represent petroleum hydrocarbon. As seen in the mineralization reactions in Equations (1), (2), and (3) below, three different compounds are typically chosen to represent total petroleum hydrocarbon (TPH): benzene (C6H6), decane (C10H22), and octane (C8H18) which was used to calculate the rate for the Shell Carson results. 2C6H6 + 15O2 12CO2 + 6H2O (1) 2C10H22 + 31O2 22H2O + 20CO2 (2) 2C8H18 + 25O2 18H2O + 16CO2 (3) The selection of this representative compound can slightly affect the total estimated biodegradation rate at a site, either expressed in units of gal/acre/yr (Table 4). The values are compared to those obtained using octane as the representative compound, as it was the selected compound for this study. As seen in the table below, rates calculated using benzene are 20% lower than those of octane. Additionally, rates calculated using decane are 5% greater than those of octane. Generally, the overall effect of selected compound on the final results is relatively small. October 2015 Page 12 of 22 Table 4: Difference in Biodegradation Rates (gal/acre-yr) Rate, Rate, Rate, CO2 Flux CO2 Flux Expressed as Expressed as Expressed as (µM/m2-sec) (g CO2/m2-day) Benzene Decane Octane* (gal/acre-year) (gal/acre-year) (gal/acre-year) 1 4 530 690 660 5 19 2510 3300 3130 10 38 5010 6580 6260 % Difference, as Compared 20% 5% -to Octane *= Octane was used in the rate calculations for the Shell Carson results. KEY POINTS Different compounds can be used to calculate hydrocarbon biodegradation rates based on the amount of CO2 generated, but the overall effect on the final results is relatively small (5% and 20% difference). Question 4. Does LNAPL Specific Volume and Geology Affect Degradation Rates? At the Shell Carson site, Rapid Optical Screening Tool (ROST) and Cone Penetrometer Testing (CPT) were applied to evaluate the extent and recoverability of LNAPL. Here, ROST data shows the fluorescence intensity, and CPT data shows geologic information at a fine-scale interval. Approximately 368 borings were acquired across the site (Geosyntec and URS, 2007). The ROST results (fluorescence) and CPT results (hydrogeology) were also used to calculate LNAPL saturations across depth-discrete intervals at all locations (Geosyntec and URS, 2007). For this analysis, the resulting ROST and CPT data from borings closest to CO2 traps were selected, as presented in Attachment 2A (LNAPL area locations) and Attachment 2B (background locations). Additionally, the calculated LNAPL saturations at depthdiscrete intervals were used to estimate a location-specific specific volume of LNAPL (by calculating the area under the depth-saturation curve). Note that there is considerable uncertainty in these specific volume estimates. As seen in Attachment 2A, locations with larger hydrocarbon degradation rates generally had larger ROST signals. This qualitative analysis was supplemented by plotting of LNAPL hydrocarbon biodegradation rate vs. specific volume at each location. As seen in Figure 5 below, there was a good correlation between specific volume and hydrocarbon degradation rates (r2 of 0.8). October 2015 Page 13 of 22 Figure 5: Correlation of the Specific Volume of LNAPL at CO2 Trap Locations to the Hydrocarbon Degradation Rates. The degradation rates were determined by 14C analysis at all locations. Additionally, water level gauging data from the carbon trap deployment time period was used to distinguish between the vadose and saturated zones at each location. The specific volume in each zone was then used to determine if the hydrocarbon biodegradation rates are related to amount of LNAPL in the vadose and saturated zones. As seen in Figure 6 and 7 below, no such correlation was found. Figure 6: Correlation of the Specific Volume of LNAPL in the Vadose Zone at CO2 Trap Locations to the Hydrocarbon Degradation Rates. October 2015 Page 14 of 22 Figure 7: Correlation of the Specific Volume of LNAPL in the Saturated Zone at CO2 Trap Locations to the Hydrocarbon Degradation Rates. Finally, the predominant geologic soil types were clay and clayey soils, with some layers of sand and silty sands. No apparent relationship was observed between geology and hydrocarbon degradation rates at the individual locations. KEY POINTS Locations with more prevalent and stronger ROST signals generally correlated with higher hydrocarbon biodegradation rates based on a qualitative study. Hydrocarbon biodegradation rates were strongly correlated to the total volume of LNAPL at each location as expressed by an LNAPL specific volume estimate derived from ROST data (r2 of 0.8), but were not correlated to the amount of LNAPL in the vadose and saturated zones. No relationship was observed between geology and hydrocarbon degradation rates. October 2015 Page 15 of 22 Question 5. How Does the Gradient Method Compare to Methods Evaluating CO2 Measured at the Surface? Two methods are commonly used to estimate hydrocarbon biodegradation rates at an LNAPL affected site, as follows: 1. Gradient Method – Involves taking depth-discrete soil gas profiles of CO2 such that concentrations are expected to increase with depth to the source zone (ITRC, 2009). The CO2 measurements are used to calculate a flux using Fick’s first law (Sihota, 2011): J Deff C s Ca z (3) Where Deff is the effective diffusion coefficient, Cs is the subsurface concentration (mol/m3), Ca is the atmospheric concentration (mol/m3), and ∆z is the soil monitoring depth in relation to the surface (Sihota, 2011). 2. Flux Chamber Method – Involves capturing and measuring CO2 flux at the surface (i.e. using carbon traps), and applying either Method 1 or Method 2 described previously, to calculate hydrocarbon biodegradation rates. CO2 flux measurements obtained using the gradient method are likely dominated by LNAPL affected areas (i.e. have high subsurface CO2 concentrations not seen in background areas). Additionally, because CO2 generated is mostly non-reactive (i.e., conservative), the flux calculated using the gradient method is basically equivalent to the flux seen at the surface. Thus, results obtained using the two methodologies should be similar. Field study results using both the gradient method and flux chamber method were compared in Sihota et al., 2011, as seen in the table excerpt below. Contaminant degradation rates using the flux chamber method (3.3 g TPH/m2-day) were found to be similar, and within the range of values obtained from the gradient method (1.1-3.6 g TPH/m2-day). (Source: Sihota, 2011) KEY POINTS Hydrocarbon degradation rates obtained from applying either the flux chamber method (i.e. capturing CO2 at the surface) are similar to those using the gradient method (i.e. taking depth-discrete subsurface CO2 measurements). October 2015 Page 16 of 22 Question 6. What are possible historic LNAPL saturation values? The measured average hydrocarbon biodegradation rate can be used to approximate the additional LNAPL volume that could have existed at the site at a specific year in the past (i.e. hydrocarbon biodegradation rate in gal/acre/yr * Y years = additional LNAPL volume that existed Y years ago). Thus, we determined the additional LNAPL volume and the estimated historic LNAPL saturation at three different years in the past. The most recent LNAPL saturation was determined to be 8.5% in the vicinity of the biosparge pilot test at the Shell Carson site, while estimated historic values range from 10%-12%. Year 1983 1963 1943 Table 5: Potential Historic LNAPL Saturation Time to Additional LNAPL Estimated Historic Present Volume LNAPL Saturation (yrs) (gal/acre) (%) 30 51,000 10 50 85,000 11 70 119,000 12 Notes: (1) Calculation assumes no additional releases occurred since listed year (2) An average hydrocarbon biodegradation rate of 1,700 gal/acre/yr used (3) Current LNAPL saturation at site is 8.5%, according to biosparge pilot test analysis KEY POINTS Approximate historic LNAPL saturation values are reasonable, further indicating that the measured hydrocarbon biodegradation rate is valid. October 2015 Page 17 of 22 ATTACHMENTS Attachment 1: CO2 Trap Locations and LNAPL Distribution, Shell Carson Site Attachment 2A: LNAPL Area Locations: Geology and LNAPL Distribution at CO2 Trap Locations Attachment 2B: Background Locations: Geology and LNAPL Distribution at CO2 Trap Locations October 2015 Page 18 of 22 REFERENCES ITRC (Interstate Technology & Regulatory Council), (2009). Evaluation of Natural Source Zone Depletion at Sites with LNAPL. Prepared for ITRC by the LNAPL Team. Bekins, B.A., Hostettler, F.D., Herkelrath, W.N., Delin, G. N., Warren, E., Essaid, H.I., 2005. Progression of Methanogenic Degradation of Crude Oil in the Subsurface. Envrionmental Geosciences 12, no. 2: 139-152. Geosyntec and URS, 2007. Light Non- Aqueous Phase Liquid (LNAPL) Investigation, Shell Oil Products Carson Terminal, Carson, California. December 2007. McCoy, Kevin M., 2012. Resolving Natural Losses of LNAPL Using CO2 Traps. Colorado State University, Department of Civil and Environmental Engineering, Master’s Thesis. Molins, S., Mayer, K.U., Amos, R.T., Bekins, B.A., 2010. Vadose Zone Attenuation of Organic Compounds at a Crude Oil Spill Site – Interactions between Biogeochemical Reactions and Multicomponent Gas Transport. Journal of Contaminant Hydrology. 112, 15-29. NRC (National Research Council: Committee on Intrinsic Remediation, Water Science and Technology Board, Board on Radioactive Waste Management). 2000. Natural Attenuation for Groundwater Remediation. Washington, D.C.: National Academy Press. Sihota, N., Singurindy, O., Mayer, U., 2011. CO2-Efflux Measurements for Evaluating Source Zone Natural Attenuation Rates in a Petroleum Hydrocarbon Contaminated Aquifer. Envrion. Sci. Technol. 45, 482-488. Zimbron, J., 2012. Basis for 14C Analysis to Differentiate from Petroleum Hydrocarbon and Modern Carbon using the CO2 Flux Traps. Colorado State University. July 2012. October 2015 Page 19 of 22 ATTACHMENT 1 CO2 Trap Locations and LNAPL Distribution, Shell Carson Site October 2015 Page 20 of 22 ATTACHMENT 2A LNAPL AREA LOCATIONS: Geology and LNAPL Distribution at CO2 Trap Locations Sorted With Hydrocarbon Degradation Rate From Highest to Lowest October 2015 Page 21 of 22 ATTACHMENT 2A LNAPL AREA LOCATIONS: Geology and LNAPL Distribution at CO2 Trap Locations Sorted With Hydrocarbon Degradation Rate From Highest to Lowest October 2015 Page 22 of 22 ATTACHMENT 2A LNAPL AREA LOCATIONS: Geology and LNAPL Distribution at CO2 Trap Locations Sorted With Hydrocarbon Degradation Rate From Highest to Lowest October 2015 Page 23 of 23 ATTACHMENT 2B BACKGROUND LOCATIONS: Geology and LNAPL Distribution at CO2 Trap Locations Memo 2: Natural Source Zone Depletion (NSZD) Results of Pilot Test at Tesoro Hynes Facility For Western States Petroleum Association October 2015 These memos were prepared by the staff of GSI Environmental, Inc., and AECOM on behalf of Western States Petroleum Association (WSPA) and the LA LNAPL Workgroup. The memos were prepared by/under the supervision of the engineers and geologists whose signatures appear below. This document is intended for the sole use and benefit of WSPA and its member industry companies. The scope of services performed as described in this memo may not be appropriate to satisfy the needs of other users, and any use or reuse of this document or the findings, conclusions, or recommendations presented herein is at the sole risk of said user. No express or implied representation or warranty is included or intended except that the work was performed within the limits prescribed by the Client with the customary thoroughness and competence of professionals working in the same area on similar projects. The findings or professional opinions were prepared in accordance with generally accepted professional engineering and geologic practice, and no warranty is expressed or implied. GSI Environmental, Inc. AECOM Environmental Engineer Frank Muramoto, P.G., C.HG Senior Program Manager /^^S^^^ October 2015 Page 1 of 12 MEMORANDUM TO: LA LNAPL Workgroup FROM: Poonam Kulkarni, Charles Newell, GSI Environmental Inc; Faraneh Chamran, Jun Lu, Daniel Buckley, Frank Muramoto, AECOM; Julio Zimbron, Colorado State University RE: Natural Source Zone Depletion (NSZD) Results of Pilot Test at Tesoro East Hynes Facility EXECUTIVE SUMMARY Carbon traps were deployed at the Tesoro East Hynes facility in order to determine the rates of hydrocarbon biodegradation at the site. Additionally, the relationship between hydrocarbon biodegradation rate and petroleum hydrocarbon impacts at the site was evaluated using relevant site data. Data reviewed included: 1) GC/FID hydrocarbon characterization of light, non-aqueous phase liquid (LNAPL) collected from the monitoring wells, and 2) Rapid Optical Screening Tool (ROST) profiles of nearby locations. Results of this pilot test and subsequent analyses are as follows: NSZD is occurring at the Tesoro East Hynes facility, and fossil fuel carbon sources. The site average hydrocarbon (LNAPL) biodegradation rate is approximately 1,100 gallons/acre/year equivalent. Hydrocarbon biodegradation rates from the pilot test (1,100 gallons/acre/year [gal/acre/yr]) are also comparable to those at six other field sites measured by CSU (McCoy, 2012), where the average LNAPL loss rate is 3,500 gal/acre/yr, and comparable to the LA LNAPL NSZD test at the Shell facility, where an average of 1,700 gal/acre/yr was measured. Hydrocarbon degradation rates do not seem to correlate with the vertical distribution of hydrocarbons, according to ROST data. Compound categories (i.e., LNAPL composition) do not appear to have any correlation with CO2, probably due to factors such as non-quantitative nature of the data and complicated microbial environment. 14 C analysis confirmed October 2015 Page 2 of 12 MEMORANDUM OBJECTIVE The LA LNAPL Workgroup, working with Colorado State University (CSU) and AECOM, deployed carbon traps at the Tesoro East Hynes facility with the goal of estimating the carbon flux at the soil surface and by extension, estimating the rates of hydrocarbon biodegradation at the Tesoro East Hynes site. The results from the carbon trap pilot test at the Tesoro East site are discussed below. BACKGROUND As previously described in the memo “Natural Source Zone Depletion (NSZD) Lines of Evidence and Results of Pilot Test at Shell Carson Facility,” CO2 generation above background levels is direct evidence of biodegradation of hydrocarbons and an accepted indicator of NSZD. Here, two methods can be used to estimate the hydrocarbon biodegradation rates at a site: 1) Method 1: LNAPL Area vs. Background Area, and 2) Method 2: 14C Analysis. At the Shell Carson Facility, both methods were employed, and the average hydrocarbon degradation rates calculated using both methods were the same (1,700 gal/acre/yr). At the Tesoro East Hynes facility, due to uncertainty about background locations (i.e., those without LNAPL), only Method 2 was employed. RESULTS Average Degradation Rates at Site A map showing the trap locations can be found in Attachment 1, and the hydrocarbon biodegradation rates determined from the 14C analysis are shown in Table 1 below. The average hydrocarbon biodegradation rate at the site was found to be 1,100 gal/acre/yr by conducting 14C analysis on all carbon traps. As seen in Table 1 below, the hydrocarbon degradation rates range from 120 to 2,660 at the site. October 2015 Page 3 of 12 Table 1: Summary of CO2 Trap Results at Shell Carson Site (Method 2: 14C Analysis) Location 1 2 3 4 5 6 7 Average Hydrocarbon Degradation Rate (gal/acre/yr equiv.)* 240 470 430 1,170 1,990 120 2,660 1,100 Notes: (1) Map of trap locations is provided in Attachment 1 (*) NSZD rates calculated using octane as the mineralization compound. KEY POINTS The site average hydrocarbon (LNAPL) biodegradation rate is approximately 1,100 gallons/acre/year equivalent. NSZD is occurring at the Tesoro East Hynes facility, and fuel carbon sources. Hydrocarbon biodegradation rates from the pilot test (1,100 gal/acre/yr) are also comparable to those at six other field sites measured by CSU (McCoy, 2012), where the average LNAPL loss rate is 3,500 gal/acre/yr, and comparable to the LA LNAPL NSZD test at the Shell facility, where an average of 1,700 gal/acre/yr was measured. 14 C analysis confirmed fossil Relationship between Degradation Rates and Petroleum Hydrocarbon Impacts at Site To understand the relationship between CO2 flux (i.e., hydrocarbon biodegradation rate) and petroleum hydrocarbon impacts at the site, relevant site data were collected and reviewed. Data reviewed included: GC/FID hydrocarbon characterization of LNAPL collected from the monitoring wells, and ROST profiles of nearby locations. 1. LNAPL GC/FID Hydrocarbon Concentrations Available GC/FID concentration data were also reviewed. Concentrations of compounds in terms of relative percentage in the LNAPL were used for correlation evaluation. Eighty-nine available compounds were grouped into five categories: 1) paraffins (i.e., straight chain saturated hydrocarbons); 2) isoparaffins (i.e., branched saturated hydrocarbons); 3) aromatics (i.e., compounds with one or more benzene rings); 4) naphthenes (i.e., saturated cyclic hydrocarbons) and 5) olefins (i.e.; unsaturated hydrocarbons excluding aromatics). Rationale for this grouping is that certain groups of compounds, especially simple aromatics and paraffins, are more sensitive to biodegradation and, therefore, the relative concentration of each category of compounds in the LNAPL may be related to the October 2015 Page 4 of 12 hydrocarbon degradation rate. The data were reviewed and no correlation was observed. 2. ROST Profiles The rapid optical screening tool (ROST) uses a tunable laser at a specific frequency passed through a fiber optic cable and emitted through a sapphire window on a standard cone penetrometer testing (CPT) system or percussion system (e.g., Geoprobe). The light causes the polycyclic aromatic hydrocarbon (PAH) constituents in the soil to fluoresce. The emitted fluorescence is transmitted via a second fiber optic cable to a measurement system within the truck. ROST data are typically presented as percent total fluorescence with depth, as shown in Attachment 2. The horizontal axis is fluorescence intensity (FI) relative to a reference (i.e., instrument standard for calibration) and the vertical axis is the depth from ground surface. To quantify the relationship between hydrocarbon degradation rate and amount of LNAPL in both the vadose and saturated zones, raw ROST data were used. First, gauging data of LNAPL and water levels during the deployment time period of the carbon traps were reviewed. These data were used to estimate the water levels, and distinction between the vadose and saturated zones in the nearest ROST location (see Figure 2). Across the soil column, ROST fluorescence collection frequency varies; but the average is 10 FI measurements per foot. Thus the total FI is integrated for the each of zones, as seen in Figure 2. Ultimately, the total integrated ROST signal was calculated for each zone. Finally, each ROST horizontal location is designated with a letter A, B or C. These letter designations indicate the relative horizontal distance from ROST locations and their respectively CO2 trap, with A being closer (approximately 50 feet or less), B being a moderate distance away (approximately 50 to 100 feet),, and C being farther than 100 feet between ROST and trap. October 2015 Page 5 of 12 Vadose Zone Saturated Zone Figure 2: Example ROST Signal and Distribution of Zones Figures 3 through 5 below show the relationship between hydrocarbon degradation rates and the total integrated ROST signal at each location, and in each zone. These figures do not show a correlation to the distribution of LNAPL in the subsurface. October 2015 Page 6 of 12 Figure 3: Relationship between Hydrocarbon Degradation Rate and Total Integrated ROST Signal. Trend not statistically significant (p=0.7). Figure 4: Relationship between Hydrocarbon Degradation Rate and Total Integrated ROST Signal in the Vadose Zone. Trend not statistically significant (p=0.2). October 2015 Page 7 of 12 Figure 5: Relationship between Hydrocarbon Degradation Rate and Total Integrated ROST Signal in the Saturated Zone. Trend not statistically significant (p=0.9). It should be pointed out that the ROST FI is related to PAHs in the LNAPL. Older gasoline typically contains significant amounts of heavier components, which can include PAHs. As gasoline weathers, as is likely the case at this site, the concentrations of heavier components are relatively enhanced, leading to a higher FI response. KEY POINTS Hydrocarbon degradation rates do not seem to correlate with the vertical distribution of hydrocarbons, according to ROST data. GC/FID compound categories do not appear to have any correlation with CO2, probably due to factors such as non-quantitative nature of the data and complicated microbial environment. ATTACHMENTS Attachment 1: Facility CO2 Trap Locations and LNAPL Distribution, Tesoro East Hynes Attachment 2: ROST Data and Hydrocarbon Degradation Rates at CO2 Trap Locations October 2015 Page 8 of 12 REFERENCES McCoy, Kevin M., 2012. Resolving Natural Losses of LNAPL Using CO2 Traps. Colorado State University, Department of Civil and Environmental Engineering, Master’s Thesis. October 2015 Page 9 of 12 ATTACHMENT 1 CO2 Trap Locations and LNAPL Distribution, Tesoro East Hynes Facility October 2015 Page 10 of 12 ATTACHMENT 2 ROST Data and Hydrocarbon Degradation Rates at CO2 Trap Locations October 2015 Page 11 of 12 ATTACHMENT 2 ROST Data and Hydrocarbon Degradation Rates at CO2 Trap Locations October 2015 Page 12 of 12 ATTACHMENT 2 ROST Data and Hydrocarbon Degradation Rates at CO2 Trap Locations